[extropy-chat] (health) risks and benefits along a bell curve ?

[Lee Corbin]

Ensel writes

> What I am saying is that the _reason_ I am not concerned
> is my common sense conclusion that:  If I don't see at least a few people
> (read: the beginning of the bell curve) dying very soon from cell phones,
> then presumably the effect is negligible.  If practice XYZ really was
> significantly harmful, I would at least see the statistical outliers
> kicking the bucket already ...
> 
> My question is, from a learned medical standpoint, is this sound reasoning
> ?  Why _wouldn't_ this be a true statement ?

The fact that normal distributions occur a lot in nature cannot
be used to suggest that we should usually expect one. A counter-
example that comes to mind is radiation poisoning. It's hardly
the case that following a catastrophe there is a slow *rise* in
the number of cases requiring treatment that's anything similar
to the trailing-off end.

Yes, I realize that it's the other end of the curve you're
thinking about: you're doubting the existence of a pronounced
effect of anything that doesn't show up for a long time, and I
do agree that in that case you'd be more frequently right. I can't
think off-hand of persuasive cases of phenomena that have long
incubation periods; but surely there are some.

In any case, no, I don't think that use of the normal "bell-curve"
distribution as an archetype is a very good idea.

I do, though, grant along with Spike that you have a very interesting
point here:

> If I don't see at least a few people that become dramatically
> healthier and live dramatically longer because of these practices,
> then probably the effect is negligible.  If practice XYZ really
> was significantly healthful, I would at least see the statistical
> outliers living to be 150 (and since the prescription is for things
> like green tea and cocoa and fruits and fiber, we would have seen
> them throughout history)

Gee, as far as I know (which isn't far), that does make a lot of
sense!

But any *good* mathematical argument to this effect has to be a lot
more sophisticated than simple appeal to a few usual distributions.

Lee